mirror of
https://github.com/saymrwulf/pytorch.git
synced 2026-05-15 21:00:47 +00:00
ed47b85d3b
Summary: so that functions like `def fn(x, p:float)` can be fused. Fixes #9940 and #11186. Fuses only float (not integer) arguments to simplify assembling arguments for fusion launch. CPU fusion is disabled in CI and this won't be tested, but I tested it locally. cc t-vi, apaszke Pull Request resolved: https://github.com/pytorch/pytorch/pull/18087 Differential Revision: D14581206 Pulled By: wanchaol fbshipit-source-id: ccb0cf79b1751706f9b2cdf1715115eae5a39fb6
29 lines
811 B
C++
29 lines
811 B
C++
#pragma once
|
|
|
|
#include <torch/csrc/WindowsTorchApiMacro.h>
|
|
#include <torch/csrc/jit/fuser/arg_spec.h>
|
|
#include <torch/csrc/jit/fuser/partition_desc.h>
|
|
#include <torch/csrc/jit/fuser/tensor_desc.h>
|
|
#include <torch/csrc/jit/ir.h>
|
|
|
|
#include <iostream>
|
|
#include <string>
|
|
#include <tuple>
|
|
#include <vector>
|
|
|
|
namespace torch {
|
|
namespace jit {
|
|
namespace fuser {
|
|
|
|
// Creates a CPU or CUDA kernel for the given graph.
|
|
// Returns the C++ or CUDA string implementing the kernel.
|
|
TORCH_API std::string generateKernel(
|
|
const std::string& name,
|
|
const Graph& graph,
|
|
const std::vector<std::pair<const Value*, const c10::optional<TensorDesc>>>& inputs,
|
|
const std::vector<std::pair<const Value*, const TensorDesc>>& outputs,
|
|
const bool use_cuda);
|
|
|
|
} // namespace fuser
|
|
} // namespace jit
|
|
} // namespace torch
|